Monitoring: Establish manual assessment process

[kartpop]

Is your feature request related to a problem?
The team lacks a monitoring cadence for all instances (backend, guardrails, frontend) before automation can begin, which may lead to unresolved issues.

Describe the solution you'd like

• Phase 1: Implement a manual monitoring pass every Wednesday morning to assess memory spikes and usage trend changes, creating a checklist during the process.
• Phase 2: Plan for automation at a later stage.
• Start a Kaapi-channel thread and draft the monitoring document.

Original issue

Context

The team needs a monitoring cadence across all instances (backend, guardrails, frontend) before automating.

Decision

• Phase 1: manual pass every Wednesday morning over all instances (memory spikes, usage trend changes), building a checklist as they go.
• Phase 2: automation, later.

Next steps

• Start a Kaapi-channel thread and drafts the doc monitoring doc.